7/26/2019 Meniare Coc 2
1/24
J Am Acad Audiol 17:4568 (2006)
45
*Hearing and Speech Department, University of Kansas Medical Center; Department of Communication Science and
Disorders, University of Pittsburgh
John A. Ferraro, Ph.D., Hearing and Speech Department, University of Kansas Medical Center, 39th and Rainbow Blvd.,
Kansas City, KS 66160-7605; Phone: 913-588-5937; Fax: 913-588-5923; E-mail: [email protected]
Electrocochleography in the Evaluation of
Patients with Mnires Disease/
Endolymphatic Hydrops
John A. Ferraro*John D. Durrant
Abstract
Electrocochleography (ECochG) has evolved as an important tool in the
diagnosis/assessment/monitoring of Mnires disease/endolymphatic hydrops
(MD/ELH).This manuscript provides an update on the use of ECochG for these
purposes.The material presented includes descriptions of the components of
the electrocochleogram;ECochG recording approaches and parameters; how
to prepare for an exam, including subject/patient considerations;construction
and placement of a tympanic membrane recording electrode;and interpretation
the electrocochleogram. Various approaches aimed at improving ECochGs
sensitivity and specificity to MD/ELH also are described. These approachesgo beyond simple measurement of the now-conventional summating potential
(SP)/action potential (AP) magnitude ratio to include the SP magnitude to
tonebursts, the SP/AP area ratio, and the AP latency difference to clicks of
opposing polarity.
Key Words: Action potential, auditory brainstem response, auditory evoked
potentials, broadband click, cochlear microphonic, electrocochleography,
extratympanic, Mnires disease/endolymphatic hydrops, summating potential,
toneburst, transtympanic, tympanic membrane, tymptrode
Abbreviations: ABR = auditory brainstem response; AC = alternating current;
AP = action potential; AEP = auditory evoked potential;BBC = broadband click;
CM = cochlear microphonic; ECochG = electrocochleography; ET =
extratympanic; MD/ELH = Mnires disease/endolymphatic hydrops; SP =summating potential; TM = tympanic membrane;TT = transtympanic
Sumario
La electrococleografa (ECochG) ha evolucionado como una importante her-
ramienta en el diagnstico, evaluacin y monitoreo de la enfermedad de
Mnire-hydrops endolinftico (MD/HEL). Este manuscrito aporta una actu-
alizacin en el uso de la ECochG para estos propsitos.El material presentado
incluye descripcin de los componentes del electrococleograma;parmetros
y enfoques de registro del ECochG; cmo prepararse para un examen,
incluyendo consideraciones para el sujeto/paciente; construccin y colocacin
de un electrodo de registro en la membrana timpnica, e interpretacin del
electrococleograma.Se describen tambin varios enfoques orientados a mejo-rar la sensibilidad y la especificidad de la ECochG en la MD/HEL. Estos
enfoques van ms all de la simple medicin de la convencional tasa de mag-
nitud del potencia de suma (SP)/potencial de accin (AP), e incluyen la magnitud
del SP para bursts tonales, la tasa de rea de SP/AP y la diferencia de laten-
cia del AP a clics de polaridad opuesta.
7/26/2019 Meniare Coc 2
2/24
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
46
As the term implies,Electrocochleography
(ECochG) is a technique for recordingthe electrical events of the cochlea.
The clinical application of ECochG, however,
is confined to the stimulus-related cochlear
potentials and often includes measurement
of the whole nerve or compound action
potential (AP) of the auditory nerve.As shown
in Figure 1, an electrocochleogram may
comprise the cochlear microphonic (CM),
cochlear summating potential (SP), and AP,
recorded independently or in various
combinations. Although the tracings in this
figure were recorded in response to broadband
clicks, tonal stimuli also are commonly used
to evoke the components of interest.
ECochG emerged as a clinical tool in the
1970s,even though attempts to record the CM
from humans date back almost to the time of
its discovery in the cat by Wever and Bray
(1930) (e.g.,Fromm et al, 1935;Andreev et al,
1939; Perlman and Case,1941;Lempert et al,
1947; Lempert et al, 1950). The SP was
described in animals in 1950 (Davis et al,
1950; von Bekesy, 1950) but received little to
no attention in humans until much later
(e.g., Eggermont, 1976; Gibson et al, 1977;
Gibson, 1978).The first recordings of human
auditory nerve APs are credited to Ruben
and his colleagues (1960), who performed
their measurements on patients undergoing
otologic surgery.
Increased attention to all auditory evoked
potentials (AEPs) began to occur in the early
1970s, following the discovery and clinicalapplication of the auditory brainstem
response (ABR). The development and
refinement of noninvasive recording
techniques also facilitated the clinical
application of AEPs, including ECochG.
The technical capability to record cochlear
and auditory nerve potentials in humans has
led to a variety of clinical applications for
ECochG, chief among them being a tool in
the diagnosis/assessment/ monitoring of
Mnires disease/ endolymphatic hydrops
Palabras Clave: Potencial de accin, respuesta del tallo cerebral, potenciales
evocados auditivos, cl ic de banda ancha, microfnica coclear,
electrococleografa, extra-timpnico, enfermedad de Mnire/hydrops
endolinftico, potencial de suma, burst tonal, trans-timpnico, membrana
timpnica, electrodo para-timpnico
Abreviaturas: ABR = respuesta auditiva del tallo cerebral;AC = corriente alterna;
AP = potencial de accin;AEP = potencial evocado auditivo;BBC = clic de banda
ancha, CM = microfnica coclear; ECochG = electrococleografa;ET = extra-
timpnico; MD/ELH = enfermedad de Mnire/hydrops endolinftico; SP =potencial de suma; TM = membrana timpnica; TT = trans-timpnico
Figure 1. Electrocochleograms evoked by broad-
band click stimuli.Top tracings show responses to rar-
efaction (R) and condensation (C) polarity clicks.
Adding R and C responses (middle tracing) enhances
the Summating Potential (SP) and auditory nerve
Action Potential (AP), which are not phase-locked to
the stimulus. Subtracting R and C responses (bottom
tracing) enhances the Cochlear Microphonic (CM)
(from American Speech-Language-Hearing Associa-
tion, 1988, p. 9, based on data from Coats, 1981).
7/26/2019 Meniare Coc 2
3/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
47
(MD/ELH) and the assessment/monitoring
of treatment strategies for these disorders.The
material for this article has been organized to
include brief descriptions of the salient
features of the components of an
electrocochleogram, and how ECochG is
recorded, interpreted and used clinically,
especially with reference to the aboveapplication. Much of this information is
derived from recent chapters by the authors
on ECochG (Ferraro, 2000; Ferraro and
Durrant,2002), and the reader is referred to
these chapters for supplemental reading in
this area.
ECOCHG COMPONENTS
Detailed descriptions of the CM, SP, andAP are abundant in the hearing scienceliterature and beyond the scope of this article.
However, be assured that the practice of
performing ECochG should not be attempted
without a thorough and working knowledge
of auditory electrophysiology. The following
section summarizes the salient features of the
CM, SP, and AP, especially as related to
clinical ECochG.
CM
The CM is an alternating current (AC)
voltage that reflects the instantaneous
displacement of the basilar membrane along
some distance within the cochlea (Ferraro
and Durrant, 2002).This distance is defined
by the effective site and method of the
recording,and the conditions of the stimulus.
As reflected by the CM,the organ of Corti acts
(in a limited way) as a microphone, but here
the tranducers are numerous as each hair cell
produces a receptor potential that is
substantially AC (but not exclusively; see
section on SP below). In animals, the CM is
perhaps the most thoroughly investigated
electrical potential of the inner ear, but itcertainly remains to be fully understood.The
historical popularity of the CM in the
laboratory derives from its link to cochlear
transduction, from well-demonstrated
sensitivity to the health of the cochlear
partition, and, certainly, because it can be
recorded from within or near the cochlea.
This latter factor is facilitated by the CMs
considerable magnitude compared to other
electrical phenomena associated with the
auditory periphery. One would think that
such features would render the CM an ideal
tool for human clinical applications. In reality,
however, the utility of the CM in differential
diagnosis of inner ear versus auditory nerve
disorders has yet to be established.Although
reductions in CM magnitude have been
reported for various disorders such as
MD/ELH (Gibson and Beagley, 1976), thesefeatures tend to reflect general rather than
specific cochlear pathology. Furthermore,
examination of the CM with confidence that
the recording represents the true potential
remains challenging in the clinical setting.
Since the CM mimics the waveform of the
evoking signal (just as the voltage output of
a microphone), it is difficult to separate from
stimulus artifact.
SP
The SP is a complex response comprising
several components. Like the CM, the SP is
stimulus related,generated by the hair cells
of the organ of Corti, and a reflection of the
displacement-time pattern of the cochlear
partition.However, whereas the CM mirrors
the stimulus waveform (i.e., time history),
the SP displays a rectified, direct current
(DC) version of this pattern more
representative of the stimulus envelope
(Dallos, 1973). The SP appears as a
unidirectional shift in the CM baseline, the
polarity of which is dictated by an interactive
effect between stimulus parameters (i.e.,
frequency and intensity) and the location of
the recording electrode.When recorded from
the tympanic membrane (TM) or ear canal,
the SP is often seen as a downward (negative)
deflection persisting for the duration of the
acoustic stimulus (see Figure 1).
Because of its complexity, the role of the
SP in hearing function remains unclear. As
DC responses to AC stimuli, however, at least
some of its components are thought to
represent nonlinearities associated with thetransduction processes in the cochlea (Tasaki
et al, 1954;Whitfield and Ross, 1965; Davis,
1968; Engebretson and Eldridge, 1968; Dallos
et al, 1972; Gulick et al, 1989; Ruth, 1994).
Whether or not the SP actually reflected
intracellular receptor potentials was
uncertain for a period of time following its
discovery. However, SP-like potentials have
since been observed inside hair cells, and it
is now clear and that the SP is not an
epiphenomenon (e.g., Dallos, 1973). It also
7/26/2019 Meniare Coc 2
4/24
48
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
has long been known that the SP is also
sensitive to mechanical and electrical biasing
(Durrant and Dallos, 1972; Durrant and Gans,
1977). The nonlinear nature of the SP has
made it useful for monitoring certain clinical
conditions such as MD/ELH, which may
augment nonlinearities in the transduction
process. This application will be discussedlater.
AP
The AP recorded via ECochG represents
the summed response of numerous, at times
thousands of, auditory nerve fibers firing
synchronously. When evoked by click stimuli,
the term whole nerve AP is applied since,
theoretically, the click has a nearly flat
spectrum over the frequency range of interest
and thus vibrates essentially the entirebasilar membrane. As recorded clinically,
however, and regardless of the stimulus, the
AP is clearly a compound action potential
-the response of a population of neurons
rather than a single unit. A stimulus with a
narrower bandwidth, such as a toneburst,
excites a more limited segment of the
membrane and, consequently, a more
restricted population of nerve fibers. Here,
then, the term whole-nerve can be
misleading. More important is the fact that
these very different stimuliclicks and
toneburstsfail to achieve their respective
objectives.That is, the spectrum of the click
that actually reaches the cochlea generally is
far from flat, due to combined earphone, ear
canal , and middle ear response
characteristics. Likewise, the cochlear
response to tonebursts is far from discrete,
due to their spectra and limited cochlear
resolution (Durrant,1986).Thus,clicks do not
excite the whole nerve,and even tonebursts
excite several points of vibration along the
basilar membrane.
A high degree of synchrony of neural
firings is essential to producing a well-defined
AP, which accounts for the popularity of click
and brief/abrupt-onset tonebursts as evoking
stimuli. In either case, however, the response
to moderately intense stimulation (i.e., 70
dB nHL or more) tends to be dominated by
neural contributions from the basal or high-
frequency end of the cochlea (Kiang, 1965),
at least in normal ears and pathological ears
with no worse than moderate hearing loss.
Since the velocity of the traveling wave is the
highest in this region, phase shifts caused by
cochlear mechanics are minimal.
The AP, like the CM, is an AC voltage.
However, unlike either of the cochlear
potentials whose waveforms reflect the
displacement-time pattern of the cochlear
partition (i.e., the CM and SP), the AP
waveform is characterized by a series of brief,predominantly negative peaks representative
of the distribution of underlying neural
firings. At suprathreshold stimulus levels,
the first and largest of these peaks is referred
to as N1.N1 is virtually the same component
as wave I of the ABR and, as such,arises from
the distal portion of the auditory nerve
(Moller and Janetta,1983).AP peaks beyond
N1 (such as N2 and N3) are analogous to
corresponding ABR components (i.e., waves
II and III) but have received little if any
clinical attention in ECochG.For clinical purposes, AP magnitude and
latency appear to be the most useful features.
The former is a reflection of the number of
nerve fibers firing. Since the afferent fibers
of the auditory nerve primarily innervate
the inner hair cells (Spoendlin, 1966), AP
magnitude also can be viewed as a reflection
of inner hair cell output. AP latency , which
is analogous to the absolute latency for
ABR components, repr esents the time
between stimulus onset and the peak of N1.
This value incorporates stimulus travel time
from the output of the transducer to the inner
ear, traveling wave propagation time along
the basilar membrane and time consumed
activating synaptic transmission between
hair cells and first order neurons.As with all
waves of the ABR, reductions in signal
intensity at suprathreshold levels for the AP
are accompanied by absolute latency
prolongations and reductions in N1
magnitude leading to eventual disappearance
into the electrical noise floor.
Since its initial recording in humans in
1960, the AP has been the most widely
studied product of ECochG. Early interest
in the AP, however, was directed toward the
development of an electrophysiological index
of hearing status in children (Cullen et al.,
1972). This effort was overshadowed by the
ABR for such purposes, primarily because
wave V of the ABR appeared to be more
sensitive and easier to measure than the AP-
N1.As AEP applications and technology have
evolved over the years, the use of the AP to
assess and monitor cochlear and auditory
7/26/2019 Meniare Coc 2
5/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
49
nerve function has received renewed
attention, especially in surgical settings. In
addition, the use of a combined AP-ABR
approach for assessing retrocochlear status
in hard-of-hearing subjects is gaining
popularity. Finally, an important application
of the AP that will be discussed in more detail
later in this paper involves the measurementof its magnitude in comparison to that of the
SP in patients suspected of having MD/ELH.
ECOCHG RECORDING APPROACHES
Transtympanic versus Extratympanic
ECochG
The terms transtympanic (TT) and
extratympanic(ET) refer to the two general
approaches for recording ECochG.TT ECochGis an invasive procedure that involves passing
a needle electrode through the TM to rest on
the cochlear promontory. A ball electrode on
the round window can also be used when the
middle ear space is exposed during surgery.
TT approaches to ECochG were introduced in
the late 1960s (e.g., Yoshie et al, 1967;Aran
and LeBert, 1968) and are still used widely
in countries other than the United States.
ET recordings are performed with an
electrode resting against the skin of the ear
canal or surface of the TM. For the latter
site, the procedure may be referred to as
tympanic (or TM) ECochG (Ferraro and
Ferguson, 1989), even though this approach
is still considered to be ET. Pioneering work
in ET recordings was performed by Sohmerand Feinmesser (1967), Coats and Dickey
(1970), and Cullen et al (1972), among others.
Although ET ECochG can be performed using
a needle electrode in the skin of the ear canal,
this option is rarely chosen. Therefore,
virtually all ET recordings are noninvasive
and,by virtue of this feature, have been better
accepted in the United States than TT
techniques.Examples of currently popular ET
electrodes are shown in Figure 2.
Both TT and ET approaches to ECochG
have advantages and disadvantages. Theprimary advantage of the TT approach is the
close proximity of the recording electrode to
the response generators, which produces
components of large magnitude with
relatively little signal averaging. The major
limitations of TT ECochG relate to its
invasiveness. Such procedures must be
performed by or require the assistance of a
physician and are therefore limited to a
medical setting.In addition, penetrating the
TM with a needle is painful to the patient,
even when local anesthetics are used.These
disadvantages certainly have limited the use
of TT ECochG in the United States.
By comparison, ECochG responses
recorded from ET sites require more signal
averaging and tend to yield smaller
component magnitudes than TT recordings.
The biggest advantage of ET approaches,
thus, is that they can be performed in
nonmedical settings with minimal discomfort
to the patient, obviating the need for
sedation/local anesthesiaand a physician.
Another factor that has facilitated the use of
ET ECochG relates to advances in electrode
design (discussed in following section) and the
practice of using the TM as a recording site.
The TM offers a good and practical
compromise between ear canal and TT
placements with respect to component
magnitudes and, consequently, signal
averaging time (Ruth and Lambert, 1989;
Ferraro, Thedinger, et al, 1994; Ferraro,
Blackwell, et al, 1994; Schoonhoven et al,
1995). Perhaps most importantly for clinical
purposes, however, the waveform patterns
Figure 2. Photographs of extratympanic ECochG
electrodes. Tymptrode (modified version of electrode
described by Stypulkoswski and Staller [1987]) (top
left), Lilly wick electrode (top right), and Bio-Logic
ECochGtrode (middle) are placed at the surface of the
tympanic membrane. Gold-foil TIPtrode (bottom)
rests in the ear canal. From Ferraro (2000, p. 429).
7/26/2019 Meniare Coc 2
6/24
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
50
that lead to the interpretation of the TT
electrocochleogram tend to be preserved in
TM recordings (Ferraro, Thedinger, et al,
1994).When performed correctly,TM ECochG
should cause minimal-to-no discomfort to the
patient. However, the technique of placing an
electrode on the highly sensitive TM can
sometimes result in more patient discomfortthan is customary for other, noninvasive ET
approaches (but certainly not as much as is
usually associated with TT ECochG).
Given the advantages and disadvantages
of both approaches, the decision to perform
ET or TT ECochG often depends on the
traditional practices,personnel,and attitudes
of the clinic. Obviously, TT recordings are
dependent on the availability of a physician
who has the time and interest to perform
the examination. While a physician is not
needed for ET ECochG, placing an electrode
on the TM is certainly a more delicate
maneuver than attaching surface electrodes
to the scalp or resting them in the ear canal.
With proper instruction and materials,
however, this procedure is relatively easy to
learn and well within the scope of professional
practice for audiologists (American Speech-
Language-Hearing Association, 1990).
Unfortunately, one factor that is virtually
overlooked in the decision to perform TT or
ET ECochG is the attitude/preference of the
patient . Given the choice with an
understanding of the benefits and limitations
of each approach, which one would you choose
if you were the patient?
ECOCHG RECORDING PARAMETERS
Selection of recording parameters forECochG varies according to thecomponents of interest. Since these
components generally occur within a latencyepoch of 5 msec following stimulus onset,
they can be considered to be in the family of
early-latencyor short-latencyAEPs (Picton
et al, 1974). As members of the same family,
ECochG components and the ABR can be
recorded using similar parameters.A notable
exception occurs in the selection of the
bandpass of the preamplifier for ECochG
when the SP is of interest.That is, the filter
setting must be wide enough to accommodate
both a quasi-steady-state DC component (the
SP) and an AC component with a
fundamental frequency of approximately 1
kHz (the AP). Other differences between
ECochG and ABR recording parameters
involve the electrode array and the number
of samples to be averaged. For ECochG, the
latter is dependent on the choice of recording
approaches, with TT requiring considerably
fewer repetitions than ET. Table 1 illustrates
suitable ET (TM) protocol for recording the
SP and AP together, which often is done when
ECochG is used in the diagnosis of MD/ELH.
A description of these parameters is provided
below.
Table 1. Extratympanic (Tympanic) ECochG Recording Protocol
Electrode Array
Primary (+) Tympanic Membrane
Secondary (-) Contralateral Mastoid or Earlobe
Common Nasion
Recording Parameters
Timebase 10 milliseconds
Amplification 50,000 XAnalog Filter Bandpass 5 Hz--3000 Hz
Repetitions 750--1000
Stimuli
Type Broadband Clicks (BBC), Tonebursts (TB)
Duration of Electrical Pulse
(BBC) 100 microseconds
Frequency (TB) 1000 Hz, 2000 Hz
Envelope (TB) 2 millisecond linear rise/fall, 10 millisecond plateau
Polarity Rarefaction, Condensation (BBC); Alternating (TB)
Repetition Rate 11.3/second
Beginning Level 80 dB nHL
7/26/2019 Meniare Coc 2
7/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
51
Electrode Array
If your preference is an electrode array
that displays the AP as a downward
(negative) deflection.The primary electrode
(i.e., the electrode connected to the
+/noninverting input of the differential
preamplifier) should rest on the TM. Sites forthe secondary (-/inverting) electrode include
the vertex of the scalp, high forehead,
contralateral earlobe, or mastoid process.We
prefer the earlobe or mastoid for the location
of the secondary (-) electrode simply because
electrodes tend to be easier to attach and
secure to these sites. The nasion, ipsilateral
earlobe, or ipsilateral mastoid may serve as
sites for the electrode connected to the
common or ground input to the
preamplifier. If you prefer the AP to be
displayed as an upward deflection (such as in
the way conventional ABR components are
displayed), simply reverse the + and - inputs
to the preamplifier.
Timebase
As indicated above, ECochG components
generally occur within the first few
milliseconds after stimulus onset. For brief
transient stimuli (such as clicks), we use a
timebase (or signal averaging window) of 10
msec, which also allows for visualization of
ABR components that follow N1. For longer
duration stimuli (such as tonebursts), the
timebase should extend beyond the duration
of the stimulus envelope so that the entire
response is observable within the averaging
window (recalling that both the SP and CM
persist for the duration of the stimulus).
Amplification Factor
Am pl if ic at ion fa ct or is se lec te d to
maximize the signal-to-noise ratio for a given
recording condition. The amount needed forsuitable recordings of the SP and/or AP for
ET measurements generally ranges between
20,000 and 100,000X,whereas the factor for
TT recordings can be much lower (by 5 to 10
times). In part,selection of this parameter is
based on the level of the electrical noise floor,
which comprises several elements (i.e.,
myogenic and electroencephalographic
activity, electrical artifact from the equipment
and/or testing environment).The sensitivity
setting of the computers analog-to-digital
converter also must be taken into account.
Thus, amplification/sensitivity settings may
vary from laboratory to laboratory and also
among evoked potential units from different
manufacturers.However, the manipulation of
these variables to provide settings
appropriate to recording conditions generally
is easily accomplished. The goal here is toamplify enough to extract a good (and real)
response without triggering the artifact
rejection routine inordinately throughout the
recording.
Analog Filter Settings
Adaptation notwithstanding, the SP, as
fundamentally a DC potential, could last as
long as the stimulus of any duration. Ideally,
then, a DC recording amplifier is needed to
record this component. However, particularly
for the amount of gain needed, such amplifiers
are notoriously unstable for
electrophysiological recordings. Fortunately,
the SP, as evoked for practical/clinical
purposes, is only quasi-steady-state,
permitting the use of the AC-coupled
amplifiers typically found in commercially
manufactured AEP units. Indeed, the click
evokes a rather brief transient DC component
that is readily recorded with low-pass cutoffs
in the vicinity of 5100 Hz, depending upon
filter characteristics. Such recordings (i.e.,
using familiar ABR settings) can be
accomplished without significant distortion
of the SP-AP complex for purposes of
measuring the SP/AP magnitude ratio
(Durrant and Ferraro, 1991). However, when
recording the SP to tonebursts, this approach
creates a DC-step-like waveform, and, even
if only quasi-steady-state, the waveform is
likely to be significantly distorted unless the
low-frequency cutoff of the amplifier is low
(generally 10 Hz or below, depending on
duration of the sound and the filter
characteristics).The low-pass (or high-frequency cutoff)
setting of the filter should be set to allow
transmission of the AC components of
interest. In the case of the AP-N1, 3000 Hz
is a suitable setting as the fundamental
frequency of this component is approximately
1000 Hz. Filter settings for CM recordings
would depend on the frequencies of the
evoking stimuli (and thus the resultant
responses).
7/26/2019 Meniare Coc 2
8/24
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
52
Repetitions
The number of individual responses
needed to extract a well-defined
electrocochleogram from the background
noise generally varies with recording
conditions, and also the subjects degree of
hearing loss. The former depends on therecording approach. That is, TT recordings
require considerably fewer repetitions than
ET approaches. More repetitions may be
necessary for subjects with hearing loss than
normally hearing subjects, especially if the
loss is in the 14 kHz range. In our
experience, when sensorineural hearing loss
in these frequencies exceeds 5060 dB HL,the
use of ET ECochG in MD/ELH populations
is questionable.The basis for this statement
is that losses of this magnitude reduce the
output of the population of hair cells
contributing to the responses of interest and
render them too small for reliable ET
recordings.On the other hand, when hearing
loss precludes the identification of wave I in
the presence of wave V in the conventionally
recorded ABR, ECochG can be very useful
(Ferraro and Ferguson, 1989).
Stimuli
As mentioned earlier, the broadband click
(BBC) is a popular stimulus for short-latency
AEPs be cause it exci tes synch ro nous
discharges from a large population of neurons
to produce well-defined peaks in the response.
In addition, 100 msec is a popular choice for
the duration of the electrical pulse driving the
transducer because the first spectral null for
a click of this duration occurs at 10,000 Hz
(i.e., 1/100 msec). In reality, the frequency
range of the transducer is usually lower than
10,000 Hz and the acoustic signal receives
additional filtering by the outer and middle
ears. Thus, the spectrum of the stimulus
reaching the cochlea is far from flat, andconsiderably narrower than 10,000 Hz.
Since the duration of both the CM and SP
are stimulus dependent, the brevity of the
click makes it a less-than-ideal stimulus for
studying either of these potentials. Despite
this limitation, the use of clicks has proven
effective in evoking the SP-AP complex for
ECochG applications related to MD/ELH,
even though the duration of the SP is
abbreviated under these conditions (Durrant
and Ferraro, 1991). This feature will receive
more attention later.
Although the click continues to remain
popular, toneburst stimuli also have been
used in several ECochG studies involving
MD/ELH populations (e.g.,Levine et al,1992;
Orchik et al, 1993; Ferraro, Blackwell, et al,
1994;Ferraro,Thedinger, et al, 1994;Koyuncu
et al,1994; Margolis et al, 1995).Tone burstsprovide a higher degree of response
frequency-specificity than clicks (depending
on stimulus envelope and duration), which
can be useful for monitoring cochlear status
in progressive disorders where hearing may
not be affected at all frequencies. In addition,
the use of longer stimuli allows for better
visualization of the SP and CM (Durrant and
Ferraro, 1991).
A lack of standardization regarding
stimulus parameters presents a problem
when using tonebursts to record ECochG
components and other AEPs. Most studies
employ signals of only one or two frequencies;
stimulus envelopes are different, and there
is no standardized approach to defining
stimulus intensity. These inconsistencies
make it difficult to compare data from
different studies/clinics. For tone bursts, we
use an envelope with a linear rise-fall time
of 12 msec and a plateau of 10 msec or
longer. Shorter plateaus (e.g., 5 msec) can
sometimes be used to inhibit (but generally
not eliminate) interference by ABR
components (Levine et al, 1992).
Stimulus polarity depends on the initial
deflection of the transducer diaphragm and
is an important factor for ECochG.Presenting
clicks or tonebursts in alternating polarity
inhibits the presence of stimulus artifact and
CM, as their phases are locked to the signal.
Stimulus artifact can sometimes be large
enough to obscure early ECochG components,
and CM generally overshadows both the SP
and AP features that are problematic when
these latter two components potentials are the
components of interest.Alternating stimuluspolarity can be applied to help overcome this
problem. However, recording separate
responses to condensation and rarefaction
clicks then adding them together off-line may
be a more preferable solution, since certain
subjects with MD/ELH display abnormal
latency differences between AP-N1 latencies
to condensation versus rarefaction clicks
(Margolis and Lilly, 1989;Levine et al, 1992;
Margolis et al, 1992; Margolis et al, 1995;
Orchik et al, 1997; Sass et al, 1997).
7/26/2019 Meniare Coc 2
9/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
53
As with the majority of signal-averaged
AEPs, it is important for ECochG that the
cochlear/neural responses to one stimulus be
complete before the next stimulus is
introduced. This requirement allows for
considerable latitude in the selection of
stimulus repetition rate for click-evoked AEPs.
For ECochG, however, increasing this ratebeyond 1030/second may cause unacceptable
adaptation of the AP (Suzuki and Yamane,
1982). Click repetition rates on the order of
100/second cause extensive (nearly total)
adaptation of the AP while leaving the SP
relatively unaffected (Gibson et al, 1977;
Coats, 1981). Unfortunately, the use of such
fast rates has not proven to be very successful
in the clinic, in part because the AP
contribution is not completely eliminated and
the SP may also be reduced under extreme
conditions (e.g.,click rates greater than 90/sec)
(Harris and Dallos, 1979; Durrant, 1986). In
addition, rapid clicks presented at loud levels
tend to be very annoying for patients.
When ECochG is performed to help
diagnose MD/ELH, the stimulus should be
intense enough to evoke a well-defined SP-
AP complex. Thus, stimulus presentation
typically begins at a level near the maximum
output of the stimulus generator.
Unfortunately, as mentioned earlier, the lack
of standardization for AEP stimuli regarding
signal calibration and dB reference is true for
both tonebursts and BBCs. Common
references include dB hearing level (HL, or
hearing threshold level [HTL]), dB normal
hearing level (nHL), dB sensation level (SL),
and dB peak equivalent sound pressure level
(pe SPL). Since the latter one is determined
by matching the SPL of a transient signal to
that of a continuous sinusoid, it represents the
only physical measure of intensity of the
three common references. It may be necessary
to calibrate ECochG signals in both nHL and
pe SPL. As with conventional audiometry,
nHL values are referenced against normalsubjects. For dB pe SPL, an oscilloscope is
used to match the level of the click to that of
a continuous sinusoid (e.g., a 1000 Hz signal).
Consistent with the findings of Stapells et al
(1982), 0 dB nHL for clicks corresponds to
approximately 30 dB pe SPL.
The use of high stimulus levels raises the
question of the need for masking to assure
unilateral stimulation of the intended ear.
This concern is irrelevant for ECochG,
however, since the magnitude of any
electrophysiological response from the nontest
ear is very small. In addition, ECochG
components are generated prior to crossover
of the auditory pathway. Finally, the potential
for crossover stimulation can be further
minimized by the use of tubal insert
earphones.
A final note regarding stimuli relates tostimulus artifact,which can be quite large for
ECochG. The nature of ET (especially TM)
electrodes is that they tend to have high
impedance and are vulnerable to radiation
from the transducer and other electrical
sources in the environment. The following
factors can help to inhibit such artifact: using
a tubal insert transducer; separating the
transducer from the electrode cables as much
as possible; braiding the electrode cables;
testing subjects in a shielded sound booth
with the examiner and AEP unit located
outside of the booth; plugging the AEP unit
into an isolated socket equipped with a true-
earth ground;using a grounded cable for the
primary electrode (such cables are
commercially available);turning off the lights
in the testing room and unplugging
unnecessary electronic equipment (it also
may be necessary to turn off the lights in
the examiner room); encasing the transducer
in grounded Mu metal shielding.
PREPARING FOR AN EXAMINATION
Recording Approach
Vi rtua lly all ECoc hG reco rd ings
performed in the authors clinics/laboratories
are ET and made from the TM because of the
advantages this site offers over other locations
along the ear canal (i.e., increased component
magnitudes, more stable/repeatable
responses, and reduced testing time because
less signal averaging is needed)
(Stypulkowski and Staller, 1987; Ruth et al.,1988; Ruth and Lambert, 1989; Ferraro and
Ferguson, 1989; Ruth, 1990; Arsenault and
Benitez, 1991). Given our preference for TM
ECochG, the fol lowing information
emphasizes this particular approach.
Subject/Patient Considerations
Most patients are unfamiliar with
ECochG and therefore confused as to what it
is,why they need it, how it will be performed,
7/26/2019 Meniare Coc 2
10/24
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
54
and whether or not it will be painful. For the
lay person (i.e., most patients) the lengthy
term electrocochleography adds to this
confusion. Instructions to the patient can
begin on the way to the testing room with an
assurance that the examination is
noninvasive and painless, that the test will
take approximately one hour, and they cansleep through it if they wish. The patient
also is informed as to why their physician has
requested this examination (e.g., to help
determine if there is too much fluid in the
inner ear). Engaging patients in conversation
at this point and watching them walk also
provides some insight regarding the status of
their hearing and balance, although the
results of other auditory and vestibular
testing should be available in their files.Once
in the sound booth, the patient is placed in
a supine position on an examining bed orsemireclined in an examination chair or
recliner. Good and comfortable head and neck
support is particularly important.Eyeglasses
and/or earrings are removed (usually by the
patient), and food/chewing gum/candy/etc.
must be swallowed or discarded. When the
patient is comfortable and attentive, he or she
is informed that devices called electrodes
will be attached to the scalp, a small, sponge-
or cotton-tipped electrode will be inserted
along the ear canal to rest on the TM,and an
earplug will be used to hold the electrode in
place and deliver click and beeping-type
sounds to the ear. The patient should be
alerted that the TM electrode might feel
strange and maybe a little uncomfortable,
but that it should not be particularly painful.
If pain should occur, repositioning the
electrode usually helps.In very rare instances
(i.e., less than 0.5% in the first authors
experience), the exam may have to be
terminated, or an ear canal electrode used
instead of the tymptrode. The procedures for
preparing the skin and placing the surface
electrodes are identical to those used for
conventional ABR examinations. Prior to
inserting the tymptrode, otoscopy is
performed to assess the patency of the ear
canal and normalcy of the TM. Cerumen
removal may be necessary to visualize the TM
and clear a pathway along the ear canal large
enough for the electrode. If either the ear
canal or TM appear abnormal or damaged,
ECochG is not advisable in general, and
certainly not without consulting the patients
physician.
As with most audiometric examinations,
both ears should be tested,even if unilateral
disease is suspected. Comparison between
affected and unaffected sides in particular can
provide important diagnostic information.
We always test the affected side first in case
the patient becomes restless as the
examination progresses.
Construction and Placement of the
TM Electrode (Tymptrode)
The photograph of ET electrodes in
Figure 2 includes the tymptrode (originally
described by Stypulkowski and Staller, 1987,
and modified by Ferraro and Ferguson,1989),
the Lilly wick electrode (Lilly and Black,
1989), the TM-ECochGtrode manufactured by
Bio-Logic, and the gold-foil TIPtrode, which
has long been available for ear canalrecordings. The latter three electrodes are
commercially available.The tymptrode can be
fabricated using the store boughtmaterials
listed below (see Durrant, 1990; Ferraro,
1992; and Ferraro, 1997):
medical grade silicon (SilasticTM)
tubing (0.058 inner diameter, 0.077
outer diameter;
TeflonTM -insulated silver wire
(0 .008 bare diameter, 0 .011
insulated diameter);
a wad of cotton;
electrode gel (not paste or cream);
fine, needle-nosed forceps;
1 cc disposable tuberculin syringe
with needle;
copper microalligator clip soldered to
the end of an electrode cable.
Briefly, the procedure for constructing
the tymptrode involves cutting the wire and
tubing into segments a few centimeters longer
than the ear canal, threading the wire
through the tubing, scraping the Teflon
insulation off of both ends of the wire,hooking
one of the bared ends into the cotton wad, and
stuffing the hooked portion of the cotton back
into the tubing with fine forceps. A small
portion of the cotton extends beyond the
tubing to serve as the electrode tip. Figure 3
is a drawing of the tymptrode constructed as
described above. Tymptrodes, at this stage,
can be made and stockpiled for indefinite
periods of time. Immediately prior to use,
the cotton tip of the tymptrode is impregnated
with electrode gel using the tuberculin
syringe. We attach the microalligator clip of
7/26/2019 Meniare Coc 2
11/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
55
the electrode cable to the other, bared end of
the wire, before inserting the tymptrode.
With the test ear is facing up, the
tymptrode is inserted into the entrance of
the ear canal and gently advanced (by hand
or using the fine forceps) until the tip touches
the TM.Contact is confirmed via otoscopy and
electrophysiological monitoring. It also helps
to ask the patient when they feel that the
electrode is touching the TM. Even with an
otoscope, it is difficult to actually see the
point of contact between the tymptrode tip
and TM in most cases. However, monitoring
the electrophysiological noise floor during
electrode placement helps to achieve propercontact. The noise floor, peak-clipping, and
cyclic activity associated with an open-line
condition drop dramatically, and the baseline
EEG becomes more stable when the TM is
contacted. Repositioning and sometimes
reinsertion of the tymptrode may be necessary
to achieve proper contact. However, using
both visual and electrophysiological
monitoring provides the best opportunity for
success on the first try.
Once the tymptrode is in place, the foam
tip of the sound delivery tube is compressedand inserted into the ear canal alongside the
electrode tubing. Care must be taken to not
push the electrode further against the TM
when inserting the earplug. Although the
materials that comprise the tymptrode are
relatively soft and flexible (which allows the
tip to compress or bend at the TM rather
than penetrate the membrane), such a
condition usually causes discomfort to the
patient. Only a portion of the transducer
earplug needs to be inserted into the canal to
hold the tymptrode in place and deliver the
signal for ECochG applications.Figure 4 is a
schematic representation of the tymptrode
and sound delivery tube in place. Even with
the most delicate contact, the TM does react
somewhat in most cases by displaying a slight
blushing spot at the point of contact with
the tymptrode. In hundreds of subjects and
patients examined by the authors, this
condition has never proven to be an untoward
reaction, clears up within minutes or hours,
and may even be a useful indicator of exactly
where the electrode was situated.
INTERPRETATION OF THE
ELECTROCOCHLEOGRAM
As with most AEPs,component magnitudeand temporal features form the basesfor interpreting the electrocochleogram.
Figure 5 depicts normal TM recordings to
alternating polarity clicks (which enhances
the SP and AP at the expense of the CM).
Component magnitudes can be measured as
absolute values (defined by single points)
Figure 3. Components of the tymptrode electrode
used for surface recordings from the tympanic mem-
brane.
Figure 4. Schematic illustration of the tymptrode in
place. Modified from Ferraro (1992, p. 28).
7/26/2019 Meniare Coc 2
12/24
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
56
(left panel) or using a baseline reference
(right panel).The authors are split on which
of these two approaches is best, showing that
even authorities of similar background and
generally similar views can still have differing
opinions on how to interpret an
electrocochleogram. In this instance, rights
and wrongs are difficult to define, particularly
under the electrically noisy circumstances
of minimally/non-invasive recording
conditions in awake patients.The first author
finds the single-point calculations to be the
more useful and potentially reliable as it
minimizes the judgmental factor of defining
an unstable or electrically skewed baseline
common to ET recordings in general.
Theoretically, this approach is subject to no
more or no less noise than single points
chosen to represent the SP and APmagnitudes. However, the second author
prefers using a value representing the
average through a millisecond or two of
baseline before the response.At least for one
of the measurements, the noise can be
reduced (the average being inherently a
variance reducer). This method is equally
useful in determining the SP magnitude for
tonebursts, for the same reason in both
casesminimizing bias by some noise peak.
In the final analysis, individual circumstances
may favor one approach over the other, but
it probably is the best idea,above all else, to
try to be as consistent as possible across
measures in the normative sample and
clinical patients. Here, for simplicity, the
single-point approach is adopted, in which
case, the normal SP measured from the TM
in response to 80 dB nHL clicks may be
expected to range from 0.11.0 microvolts,
with a mean of 0.4 microvolts.AP magnitudes
can be as large as 5.0 microvolts,although our
mean value is approximately 2.0 microvolts.
AP-N1 latency is measured from stimulus
onset to the peak of N1 and, as mentioned
earlier in this manuscript,should be identical
to the absolute latency of ABR wave I. At 80
dB nHL, normal N1 latencies generally range
from 1.31.7 msec with a mean of
approximately 1.5 msec. Since a tubal insertearphone was used to deliver the stimulus,
the above latency values have been corrected
for the 0.9 msec delay attributable to the
sound tube. Although labeled in Figure 5,
N2 has received little interest for ECochG
applications.
Also as shown in Figure 5, we measure
SP and AP magnitudes from the leading edge
of both components.The resultant values are
then used to derive the SP/AP magnitude
ratio, which is a key measure when ECochG
Figure 5. ECochG recordings from the tympanic membrane to clicks presented in alternating polarity at 80dB nHL.The magnitudes of the Summating Potential (SP) and Action Potential (AP) can be measured from peak-
to-trough (left panel), or with reference to a baseline value (right panel). Magnitude/time scale is 1.25 micro-
volts/1 millisecond per gradation. Insert phone delay is 0.90 milliseconds. From Ferraro (2000, p. 435).
7/26/2019 Meniare Coc 2
13/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
57
is used to help diagnose and monitor
MD/ELH. This value tends to occur within a
relatively small range (i.e., 0.10.5), especially
in comparison to the individual magnitudes
of both the SP and AP (Ferraro and Krishnan,
1997).
Figure 6 depicts a normal waveform
evoked by an 80 dB nHL,2000 Hz toneburst
(2 msec rise/fall, 10 msec plateau, alternating
polarity). Recall that for tonebursts the SP
persists as long as the stimulus and therefore
is not seen as a small shoulder preceding the
AP (which is still seen near the onset of the
response). To minimize the influence of the
AP, SP magn it ude is me asur ed at the
midpoint of the waveform with reference to
baseline magnitude. The polarity of the SP
depends on whether this voltage is above
(positive SP) or below (negative SP) the
baseline voltage. Figure 7 illustrates
toneburst SPs at several frequencies recorded
from both the TM and promontory (TT) of thesame normally hearing subject. Three
important features should be noted from
these tracings: (1) The polarities of the SPs
may vary slightly across frequencies. (2)
Despite these slight variations, the
magnitudes of toneburst-SPs in normal ears
are very small, which renders the actual
polarity of the SP in this population
somewhat inconsequential. (3) Although the
magnitudes of the TM responses are
approximately one-quarter that of the
promontory responses (note magnitude
scales), the corresponding patterns of the TM
and TT recordings at each frequency are
virtually identical (as indicated earlier in
this manuscript).
Although clicks and tonebursts are the
stimuli of choice for ECochG, it is important
to note that comparisons among studies from
different laboratories/clinics remain difficult
at best. This problem persists because of a
lack of consistency and standardization
regarding such aspects as choice of recording
approach,recording and stimulus parameters,
stimulus calibration, and measurement
preferences.All of these conditions continue
to necessitate the establishment of laboratory-
/clinic-specific norms for ECochG.
ECOCHGS ROLE IN THE
EVALUATION OF MD/ELH
Although much has been learned aboutMD (or idiopathic ELH) since its initial
description in the literature over 140 years
ago (Mnire,1861), the true pathophysiology
of this disorder(s) continues to be elusive.As
a result, neither a cure nor an effective
treatment strategy that works for all patients
has been developed. The symptoms upon
which diagnosis of MD/ELH is based include
recurrent, spontaneous vertigo,hearing loss,
aural fullness, and tinnitus (American
Academy of OtolaryngologyHead and Neck
Figure 6. ECochG recording from the tympanic membrane to a 2,000 Hz toneburst (alternating polarity, 80
dB nHL, 2 msec rise/fall, 10 msec plateau). Action Potential (AP) and its first negative peak (N1) are seen at
the onset of the response. Summating Potential (SP) persists as long as the stimulus. SP magnitude is meas-ured at midpoint of response (point B), with reference to a baseline value (point A). From Ferraro, Blackwell,
et al. (1994, p. 19).
7/26/2019 Meniare Coc 2
14/24
Surgery Committee on Hearing and
Equilibrium, 1995). However, the presence
and severity of these symptoms tend to vary
over time both among and within patients.
The capricious nature of this disorder makesit difficult to diagnose and evaluate with a
high degree of specificity and/or sensitivity.
As mentioned throughout this paper,
ECochG has emerged as one of the more
powerful tools in the diagnosis, assessment,
and monitoring of MD/ELH, primarily
through the measurement of the SP and AP.
In particular, it is now well documented that
the electrocochleograms of patients with
MD/ELH often display abnormally enlarged
SP magnitudes (e.g., Schmidt et al, 1974;
Gibson et al,1977;Gibson, 1978;Moriuchi andKumagami, 1979; Morrison et al,1980; Coats,
1981,1986; Kitahara et al, 1981;Goin et al.,
1982; Kumagami et al, 1982; Ferraro et al,
1983; Ferraro et al, 1985; Staller, 1986;
Dauman et al, 1988;Ruth et al, 1988;Ferraro
and Krishnan, 1997). The conventional
rationale for this finding is that an increase
in endolymph volume creates mechanical
biasing of vibration of the organ of Corti to
which,again,the SP is sensitive.Whether the
nature of this increased distortion is
mechanical (Gibson et al, 1977) and/or
electrical (Durrant and Dallos, 1972, 1974;
Durrant and Gans, 1977) has not been
resolved, and other factors such as
biochemical and/or vascular changes mayalso be responsible (Eggermont, 1976; Goin
et al; 1982; Staller, 1986). Regardless of the
specific pathophysiology, measurement of the
SP to help diagnose, assess, and monitor
MD/ELH has emerged as a primary, and
probably the most popular, application for
ECochG.
Although it is the enlargement of the SP
magnitude that often characterizes the
electrocochleograms of patients with
MD/ELH, the consistency of this finding when
using click stimuli improves when this valueis compared to the magnitude of AP-N1 to
form the SP/AP magnitude ratio (Eggermont,
1976;Coats, 1981;Coats, 1986). An enlarged
SP/AP magnitude ratio to click stimuli,
therefore, would be considered a positive
finding for ELH.This feature is illustrated in
Fig. 8, which displays the click-evoked
electrocochleogram of a patient with MD/ELH
on the left side. As can be seen from these
tracings, the left SP/AP magnitude ratio
(when measured using the single point
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
58
Figure 7. Transtympanic (TT) and Tympanic Membrane (TM) electrocochleograms evoked by tonebursts of dif-
ferent frequencies presented at 80 dB nHL (2 msec rise fall, 10 msec plateau). Stimulus frequency in kHz indi-
cated at the right of each waveform. Despite differences in magnitudes (see magnitude scale),TT and TM response
patterns are virtually identical. From Ferraro, Blackwell, et al, (1994, p. 20).
7/26/2019 Meniare Coc 2
15/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
59
method), is approximately 21/2 times larger
than the normal right ratio. Figure 9 displays
normal and abnormal toneburst-evoked
electrocochleograms (2000 Hz, 2 msec r/f, 10
msec plateau) from another MD/ELH patient.
For these measurements, SP magnitude
represents the difference between baseline and
midpoint voltages (recalling that the SP persists
for the duration of the stimulus).The left panel
displays an enlarged, negative SP from the
affected left ear, whereas the right response is
barely measurable and well within normallimits. It also should be noted for toneburst
responses that the measurement of interest is
the magnitude of the SP trough rather than the
SP/AP magnitude ratio. Indeed, it often is the
case that the AP to tonebursts may not even be
visible in the face of an abnormally enlarged SP.
Although the specificity of ECochG in
the diagnosis of MD/ELH has been reported
to be higher than 90% (Ferraro et al, 1983; Pou
et al, 1996;Murphy et al, 1997),the incidence
of an enlarged SP and SP/AP magnitude ratio
in the general Mnires population is only
approximately 55%65% and has been
reported to be as low as 20% (Gibson et al,
1977; Coats, 1981; Kitahara et al, 1981;
Kumagami et al, 1982; Margolis et al, 1995;
Pou et al, 1996; Ferraro and Tibbils, 1999).
These statistics demand a continuing search
for ways to make ECochG more sensitive,
particularly for patients whose symptoms are
not classic and for whom the clinical profile
is unclear (Campbell et al, 1992).
The episodic nature of MD/ELH certainlyplays a role in the sensitivity of any diagnostic
tool used for this disorder. In addition,
MD/ELH is characterized by various
stages during its evolution within a given
patient , which manifest differently
electrophysiologically (Aran et al, 1984;
Horner and Cazals, 1988). Thus, one might
expect the sensitivity of ECochG to vary
according to when the test was administered
in the course of the disease. In support of
this notion,Ferraro et al (1985) found positive
Figure 8. ECochG tracings to broadband clicks (alternating polarity, 80 dB nHL) from a patient with endolym-
phatic hydrops. Affected side (left) shows a magnitude-enlarged summating potential (SP), and SP/action
potential (AP) magnitude ratio. SP-AP relationship on the unaffected side (right) is within normal limits.
7/26/2019 Meniare Coc 2
16/24
electrocochleograms in over 90% of patientswho had active symptoms at the time of
testing that included aural fullness and
hearing loss. Pou et al (1996) observed
changes in the SP-AP relationship as a
function of degree of hearing loss. Such loss
may be initially episodic but also tends to
progress with duration of disease.Given these
associations, one way to make ECochG more
sensitive is to test patients when they are
experiencing symptoms. Unfortunately, the
practicality of this idea is questionable given
the fluctuating nature of the disorder(especially in its early stages), the general
operating hours of most clinics, and the
inability and/or unwillingness of patients to
complete an examination during an attack.
Nonetheless, testing patients during or as
soon as possible after an episode will
significantly improve the chances of obtaining
a positive electrocochleogram if indeed the
patient has MD/ELH. We also have found it
helpful, especially in the early stages of the
disorder, to test patients when they are
asymptomatic and then retest them whensymptoms are present . Comparing
electrocochleograms under these conditions
often reveals differences that are
diagnostically significant.
Additi on al ev idence for the cl ose
relationship between ECochG and the acute
status of the inner ear can be seen in cases
involving intraoperative ECochG monitoring
in patients undergoing endolymphatic
decompression/shunt surgery for treatment
of MD/ELH. Although still controversial,
decompression or shunting of theendolymphatic sac is an option for patients
who fail nonsurgical treatments.During such
surgeries, instantaneous measurements of
the mechanoelectrical processes of the inner
ear can be achieved via ECochG (Gibson et
al,1988;Gibson and Arenberg, 1991;Arenberg
et al,1993;Wazen, 1994; Mishler et al, 1994).
Figure 10 exemplifies intraoperative changes
in the electrocochleogram induced by probing
for the endolymphatic duct in a patient with
MD/ELH. The uppermost tracings display
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
60
Figure 9. ECochG tracings to a 2,000 Hz toneburst (alternating polarity, 80 dB nHL, 2 msec rise/fall, 10 msec
plateau) from another patient with endolymphatic hydrops. Affected side (left panel) displays an enlarged SP
trough,while SP magnitude on unaffected side (right panel) is very small and within normal limits.AP-N1 com-
ponent is seen at the onset of the response to toneburst stimuli.
7/26/2019 Meniare Coc 2
17/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
61
an enlarged SP and SP/AP magnitude ratio.
However, the SP becomes smaller and
remains that way after a metal probe is
passed into (and therefore decompresses) the
duct (bottom three tracings). Probing of
surrounding tissue did not alter the
electrocochleogram. Figure 11 displays
selected tracings measured from anotherpatient undergoing endolymphatic sac
decompression surgery.A noticeable reduction
in the SP/AP magnitude ratio to click stimuli,
and the SP magnitude to tonebursts were
observed when the sac was decompressed.
Although not always the case, this particular
patient reported an improvement in
symptoms following surgery.
Beyond the SP/AP Magnitude Ratio
Other approaches to increasing the
sensitivity of ECochG have been directed
toward the parameters associated with
recording and interpreting the
electrocochleogram, and looking beyond
comparatively simple measures of SP
magnitude. An example of such a method
involves measuring the AP-N1 latency
difference between responses to condensation
versus rarefaction clicks (as described earlier
in this chapter). Figure 12 illustrates this
procedure.The AP-N1 latency difference (LD)
between clicks of opposite polarity for this
MD/ELH patient was 0.75 msec, which was
considerably above the upper limit 0.38 msec
seen in normal ears. The basis for comparingAP-N1 latencies to clicks of opposite polarity
relates to changes in the velocity of the
traveling wave in an endolymph-loaded
cochlea. That is, the vibratory cycle of the
cochlear partition under such conditions may
be abnormally restricted (or enhanced) in
one direction over the other. If this condition
occurs, the velocity of the traveling wave (on
which the AP-N1 latency is dependent) will
differ if the initial movement of the cochlear
partition is upwards (as with rarefaction
clicks) versus downwards (as with
condensation clicks).
Another interesting feature in Figure 12
is that the AP-N1 latency difference is obscured
when responses to rarefaction and condensation
clicks are combined (lowest tracing). This
approach is analogous to presenting clicks in
alternating polarity. What appears instead is
Figure 10. ECochG recordings measured at various events during endolymphatic sac decompression surgery.
Note the reduction in SP magnitude in the Probe passed into duct tracing, which illustrates the relationship
between the electrocochleogram and the acute status of the endolymphatic system. This feature is used to help
to differentiate the location of the endolymphatic duct from surrounding tissue during surgery. From Gibson
and Arenberg (1991, p. 300).
7/26/2019 Meniare Coc 2
18/24
an abnormally widened SP-AP complex. It is
interesting to note that Morrison et al (1980)
reported a widening of the SP-AP duration in
Mnires patients over 20 years ago. Thisfinding was attributed to an after-ringing of
the CM caused by endolymphatic hydrops. In
light of recent studies, it may be more likely that
differences in AP-N1 latency to condensation
versus rarefaction clicks accounted for the
widened SP-AP complex observed by Morrison
et al (who used click stimuli presented in
alternating polarity).
Even though the underlying mechanisms
may be unclear, the above studies suggest that
the width (i.e.,duration) of the SP-AP complex
may be important to consider in theinterpretation of the electrocochleogram.
Ferraro and Tibbils (1999) explored this
notion by combining both magnitude and
duration features of the response to measure
the areas o f the SP and AP. Area
measurements were accomplished using a
special software routine that allowed us to
measure the area under the curve defined
by a straight line connecting two cursor
points. Figure 13 displays representative
tracings from this study.The waveforms in the
left panel are from a normal subject,whereas
the right tracings are from an MD/ELH
patient. The shaded portions of the top
tracings in both panels represent the area ofSP, which was defined by the onset of the SP
(baseline) and that point in the tracing where
the waveform returned to the baseline
mag ni tude. Despi te i ts l abel , t hi s
measurement also includes the areas of
components other than the SP (such as the
AP-N1, and often AP-N2) . The shaded
portions of the lower tracings represent the
AP-N1 area. The results from this study
revealed that virtually all MD/ELH patients
with enlarged SP/AP magnitude ratios also
have enlarged SP/AP area ratios. However,enlarged area ratios also were seen in several
patients suspected of having MD/ELH but
whose SP/AP magnitude ratios were within
normal limits. Subsequent research in our
laboratory using data from 138 patients with
MD/ELH has shown that measurement of
the SP/AP area ratio significantly improves
the diagnostic sensitivity of ECochG in
comparison to the SP/AP magnitude ratio
(Devaiah et al, 2003).
Given the above findings, we routinely
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
62
Figure 11. ECochG tracings recorded during endolymphatic shunt decompression surgery: baseline tracing (1),
drilling on mastoid (2), probing for endolymphatic duct (3), inserting prosthesis (4), closing (5).Tracing 5 shows
a reduction in the summating potential (SP)/action potential (AP) magnitude ratio compared to tracing 1. Trac-
ings 15 are in response to clicks, whereas tracings 67 were recorded to tonebursts at the onset of surgery and
display an enlarged SP magnitude. From Ferraro (2000, p. 446).
7/26/2019 Meniare Coc 2
19/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
63
Figure 13. Measurement of the areas of the summating potential (SP) and action potential (AP) to click stim-
uli to derive the SP/AP area ratio. Area values are obtained using a software routine that allows for measure-
ment of the area under a curve defined by a straight line that connects two cursor points. Shaded portions
represent these areas. Tracings in the top and bottom left panels are from the affected side of a subject with
endolymphatic hydrops and display an enlarged SP/AP area ratio. Normal tracings on the right are from the
unaffected side. From Ferraro and Tibbils (1999, p. 24).
Figure 12. ECochG tracings to broadband clicks presented in opposing polarity from a patient with Menieres
disease.Top tracing evoked with condensation-polarity clicks; middle tracing evoked with rarefaction-polarity
clicks.The latency difference of 0.75 milliseconds between AP-N1 components to condensation versus rarefac-
tion clicks is a positive finding for endolymphatic hydrops since it is greater than 0.38 milliseconds. This fea-
ture is obscured if the condensation and rarefaction tracings are combined to derive the response to alternat-
ing clicks (bottom tracing). From Margolis et al (1995, p. 52).
7/26/2019 Meniare Coc 2
20/24
include measurement of the SP/AP area ratio
in the interpretation of electrocochleograms.
However, we continue to measure the SP/AP
magnitude ratio to clicks, the SP magnitude
to tonebursts (1000 and 2000 Hz), and the N1
latency difference to clicks of opposite polarity.
The question as to which of these
measurements (or combination ofmeasurements) is most sensitive to MD/ELH
remains to be answered. To address this
issue, we have begun an outcome study
involving several hundred patients seen for
ECochG examinations in our clinic during the
past five years.This project involves assessing
the relationship between the results of an
ECochG examination (i.e., the individual
measurements described above) and the
subsequent diagnosis and treatment of the
patient. At this time, the jury is still out
regarding the measurement(s) that is most
sensitive to MD/ELH.
A final note regarding the specificity of
ECochG relates to enlarged SP/AP magnitude
ratios in conditions other than MD/ELH. In
particular, this feature also has been reported
for perilymphatic fistulae (Kobayashi et al,
1993;Ackley et al,1994;Campbell and Abbas,
1993, 1994).Thus, it may be the case that any
change in cochlear fluid pressure that biases
cochlear transduction,and therefore the SP,
will affect the SP/AP relationship (Storms
et al, 1996). However, as implied in earlier
discussion,definitive proof of this mechanism
remains to be seen. Eggermont (1976)
questioned the hydromechanical rationale
over two decades ago, suggest ing
consideration of possible biochemical changes
underlying or triggered by the fluid pressure
changesionic diffusion gradients can be
altered by osmosis and vice versa.
Yamasoba et al (1993) observed enlarged
SPs in cases of idiopathic low-frequency
hearing losses, namely patients who have no
balance-related or other symptoms
characteristic of MD/ELH. This findingsuggests that a positive electrocochleogram
may be a manifestation of cochlear hydrops.
Indeed, we have seen numerous patients who
f i t this pro fi le . That is, a posi tive
electrocochleogram in the face of hearing
loss, aural fullness, and tinnitus, but not
accompanied by vertigo. Likewise, we also
have observed negative electrocochleograms
in dizzy patients whose auditory symptoms
were minimal at the time of testing but who
were eventually diagnosed with MD/ELH.
Obviously, more research in this area is
needed to ascertain ECochGs sensitivity to
cochlear versus vestibular hydrops,assuming
that these conditions can exist independently.
As we continue to learn more about the
infrastructure and physiology of the inner ear,
it is likely that even more basic mechanisms
will be revealed to account for normal andimpaired hair-cell transduction in general.
Molecular biological bases, perhaps even
involving genetic coding or related processes
may play a role in altering these mechanisms
in individuals with MD/ELH to produce the
changes in the SP and other features of the
electrocochleogram described above.
SUMMARY
E
CochG has emerged as an important tool
in the diagnosis, assessment, and
monitoring of MD/ELH despite continuing
debate on the best ways to record and
interpret responses. In this case, the best
waysare those that optimize the sensitivity
and specificity of ECochG in the evaluation
of MD/ELHwhile causing minimal
discomfort to the patient. The following is a
summary of the features described in this
manuscript that should be considered in the
selection of ECochG parameters to achieve
this goal.
Althoug h ECochG compone nts
inc lude the CM, SP, and AP,
measurement of the SP and AP
continues to offer the most useful
information in the evaluation of
MD/ELH.While the CM may indeed
have other valuable applications
(e.g., in the diagnosis of auditory
neuropathy), its utility for MD/ELH
purposes has yet to be established.
The authors prefer an ET recording
approach for ECochG that involves
placing the primary electrode on the
external surface of the TM.This siteoffers a good compromise between
TT and other ET sites regarding
component magnitudes,preservation
of diagnostic patterns, and testing
time, and can be performed by
audiologists in a nonmedical setting
with minimal-to-no discomfort to the
patient.
The stimulus and signal averaging
parameters associated with ECochG
should be chosen to favor
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
64
7/26/2019 Meniare Coc 2
21/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
65
measurement of the SP and AP,
recorded together or separately, and
in response to both BBCs and
tonebursts.
The lack of standardized stimuli and
stimulus calibration approaches for
ECochG continues to necessitate the
establishment of laboratory-/clinic-specific normative data.
Stimulus and other artifact can be a
bigger problem for TM-ECochG than
for other AEPs because of the
sensitivity/fragility of recording site,
the construction, sensitivity, and
stability of the recording electrode,
and lowering the low-frequency
cutoff of the analog filter to allow
for measurement of the SP
component. Several suggestions are
offered in this manuscript to help
inhibit such artifact.
Although the SP/AP magnitude ratio
to BBCs appears to be highly specific
to MD/ELH (i.e., patients that display
an enlarged ratio are usually
diagnosed with this disorder), the
sensitivity of this measurement (i.e.,
the likelihood of finding an enlarged
SP/AP magnitude ratio in someone
who has MD/ELH) in the general
MD/ELH population is not as high.
This finding demands a continued
search for other features of the
electrocochleogram that may improve
ECochGs sensitivity.These features
may include the SP magnitude to
tonebursts, the SP/AP area ratio to
BBCs, and the latency difference
between the AP-N1 components to
BBCs of opposing polarity. Although
our data suggest that measurement
of the SP/AP area ratio may offer the
most sensitivity among the above
values,additional research is needed
to verify this finding. There is a significant relationship
between ECochG results and the
symptoms the patient displays at
the time of testing.Thus, one way to
make ECochG more sensitive is to
test patients when they are
symptomatic, even though the
practicality of this strategy often is
limited.
Although, as indicated above,
ECochG is highly specific to
MD/ELH, enlarged SP/AP
magnitude ratios also have been
reported in cases of perilymphatic
fistulae and apparent cochlear
hydrops. This finding suggests that
the conditions to which ECochG may
be most specific are changes in
cochlear fluid pressure. The relationship between ECochG
results and symptoms and also to
the acute status of the ear (as
observed in intraoperative recordings
and conditions other than MD/ELH
that cause changes in cochlear fluid
pressure), serves to underscore the
general truism of ECochG, and all
AEP measures for that matter: they
are functional indicators and not
pathognomonic of a particular
disorder/disease. Fortunately for the
diagnostician, an increased SP/AP
magnitude and/or area ratio, or
abnormally prolonged AP-N1 latency
difference to BBCs of opposing
polarity are rare in retrocochlear
cases and other disorders of hearing.
Thus, a positive electrocochleogram
certainly points strongly to a cochlear
disorder and most likely an etiology
of MD/ELH.
REFERENCES
Ackley RS,Ferraro JA,Arenberg IK. (1994) Diagnosisof patients with perilymphatic fistula.Semin Hear15:3741.
American Academy of Otolaryngology-Head andNeck Surgery. Committee on Hearing andEquilibrium. (1995) Guidelines for the diagnosis andevaluation of therapy in Menieres disease. OtolaryngolHead Neck Surg 113:181-185.
American Speech-Language-Hearing Association.(1988) The short latency auditory evoked potentials:a tutorial paper by the Working Group on AuditoryEvoked Potential Measurements of the Committeeon Audiologic Evaluation.
American Speech-Language-Hearing Association.(1990) Competencies in auditory evoked potentialmeasurement and clinical applications. WorkingGroup on Auditory Evoked Potential Measurementsof the Committee on Audiologic Evaluation. Suppl 2.
Andreev AM,Aropova AA,Gersuni SV. (1939) On elec-trical properties in the human cochlea.J PhysiolUSSR 22:206212.
Aran JM, Lebert G. (1968) Les responses nerveusecochleaires chex lhomme, image du fonctionnementde loreille et nouveau test daudiometrie objectif.Revue de Laryngologie, Otologie,Rhinologie (Bordeaux)89:361365.
7/26/2019 Meniare Coc 2
22/24
Ar an J- M, Ra re y KE , Hawk in s JE Jr. (1 98 4)Functional and morphological changes in experi-mental endolymphatic hydrops.Acta Otolaryngol97:547557.
Arenberg IK. (1980) Abnormalities, congenital abnor-malities and unusual anatomic variations of theendolymphatic sac and vestibular aqueduct: clinical,surgical and radiographic correlations.Am J Otol2:118149.
Arenberg IK, Gibson WPR, Bohlen HKH.(1993)Improvements in audiometric and electrophysiologicparameters following nondestructive inner ear sur-gery utilizing a valved shunt for hydrops andMenieres disease.Proceedings of the Sixth AnnualWorkhops on Electrocochleography & OtoacousticEmissions . Denver, CO: International MenieresDisease Research Institute, 545561.
Ar se na ul t MD, Be ni te z JT. (1 99 1)Electrocochleography: a method for making theStypulkowski-Staller electrode and testing technique.Ear Hear 12:358360.
Campbell KC,Abbas PJ. (1993) Electrocochleography
with postural changes in perilymphatic fistula andMenieres disease: case reports.J Am Acad Audiol4:376383.
Campbell KC,Abbas PJ. (1994) Electrocochleographywith postural changes in perilymphatic fistula.Animalstudies.Ann Otol Rhinol Laryngol 103:474482.
Campbel l KC, Harker LA, Abbas PJ. (1992)Interpretation of electrocochleography in Menieresdisease and normal subjects. Ann Otol RhinolLaryngol 101:496500.
Coats AC. (1981) The summating potential andMenieres disease.Arch Otolaryngol 104:199208.
Coats AC. (1986) Electrocochleography: recordingtechnique and clinical applications.Semin Hear7:247266.
Coats AC, Dickey JR. (1970) Non-surgical recordingof human auditory nerve action potentials andcochlear microphonics.Ann Otol Rhinol Laryngol29:844851.
Cullen JK, Ellis MS, Berlin CI, Lousteau RJ. (1972)Human acoustic nerve action potential recordingsfrom the tympanic membrane without anesthesia.Acta Otolaryngologica 74:1522.
Dallos P. (1973) The Auditory Periphery: Biophysicsand Physiology. New York: Academic Press.
Dallos P, Schoeny ZG, Cheatham MA. (1972) Cochlearsummating potentials: descriptive aspects.ActaOtolaryngologica 301(Suppl.):146.
Dauman R, Aran JM, Sauvage RC, Portmann M.(1988) Clinical significance of the summating poten-tial in Menieres disease.Am J Otol 9:3138.
Davis H. (1968) Mechanisms of the inner ear.AnnOtol Rhinol Laryngol 77:644656.
Davis H, Fernandez C,McAuliffe DR. (1950) The exci-tatory process in the cochlea.Proc Natl Acad Sci36:580587.
Devaiah AK, Dawson KL, Ferraro JA, Ator G. (2003)Utility of area curve ratio: electrocochleography inearly Menieres Disease.Arch Otolaryngol Head NeckSurg 129:547551.
Durrant JD. (1981) Auditory physiology and an audi-tory physiologists view of tinnitus.J Laryngol Otol4(Suppl.):2128.
Durrant JD. (1986) Combined ECochG-ABR versus
conventional ABR recordings.Semin Hear7:289305.
Durrant JD. (1990) Extratympanic electrode supportvia vented earmold.Ear Hear 11:468469.
Durrant JD, Dallos P. (1972) Influence of direct cur-rent polarization of the cochlear partition on thesummating potential.J Acoust Soc Am 52:542552.
Durrant JD, Ferraro JA. (1991) Analog model ofhuman click-elicited SP and effects of high-pass fil-tering.Ear Hear 12:144148.
Durrant JD, Gans D. (1977) Biasing of the summat-ing potentials.Acta Otolaryngologica 80:1318.
Eggermont JJ. (1976) Summating potentials in elec-trocochleography: relation to hearing disorders. In:R ub en RJ, E lb er li ng C, S al om on G, e ds.Electrocochleography. Baltimore:University ParkPress, 6787.
Engebretson AM, Eldridge DH. (1968) Model for thenonlinear characteristics of cochlear potentials.JAcoust Soc Am 44:548554.
Ferraro JA. (1992) Electrocochleography: how - partI.Audiol Today 4(6):2628.
Ferraro JA. (1997)Laboratory Exercises in AuditoryEvoked Potentials. San Diego: Singular PublishingGroup.
Ferraro JA.(2000) Electrocochleography. In: RoesserRJ, Valente M, Hosford-Dunn H, eds.Audiology:Diagnosis. New York:Thieme, 425450.
Ferraro JA, Arenberg IK, Hassanein RS. (1985)Electrocochleography and symptoms of inner ear dys-function.Arch Otolaryngol 111:7174.
Ferraro JA, Best LG, Arenberg IK. (1983) The use ofelectrocochleography in the diagnosis, assessmentand monitoring of endolymphatic hydrops. OtolaryngolClin N Am 16:6982.
Ferraro JA, Blackwell W, Mediavilla SJ,Thedinger B.(1994) Normal summating potential to tonebursts
recorded from the tympanic membrane in humans.J Am Acad Audiol 5:1723.
Ferraro JA, Durrant JD.(2002) Electrocochleography.In: Katz J, ed.Handbook of Clinical Audiology. NewYork: Lippincott, Williams and Williams, 249273.
Ferraro JA, Ferguson R. (1989) Tympanic ECochGand conventional ABR: a combined approach for theidentification of wave I and the I-V interwave inter-val.Ear Hear 3:161166.
Ferraro JA, Krishnan G. (1997) Cochlear potentialsin clinical audiology.Audiol Neurootol 2:241256.
Journal of the American Academy of Audiology/Volume 17, Number 1, 2006
66
7/26/2019 Meniare Coc 2
23/24
Electrocochleography in the Evaluation of Patients/Ferraro and Durrant
67
Ferraro JA,Thedinger B,Mediavilla SJ, Blackwell W.(1994) Human summating potential to tonebursts:observations on TM versus promontory recordings inthe same patient.J Am Acad Audiol 6:217224.
Ferraro JA,Tibbils R. (1999) SP/AP area ratio in thediagnosis of Menieres disease.Am J Audiol 8:2128.
Fromm B, Bylen CO, Zotterman Y. (1935) Studies inthe mechanisms of Wever and Bray effect. Acta
Otolaryngologica 22:477483.
Gibson WPR. (1978)Essentials of Electric ResponseAudiometry. New York: Churchill and Livingstone.
Gibson WPR,Arenberg IK. (1991) The scope of intra-operative electrocochleography. In: Arenberg IK, ed.Proceedings of the Third International Symposiumand Workshops on the Surgery of the Inner Ear.Amsterdam: Kugler Publications, 295303.
Gibson WPR, Arenberg IK, Best LG. (1988)Intraoperative electrocochleographic parameters fol-lowing nondestructive inner ear surgery utilizing avalved shunt for hydrops and Menieres disease. In:Nadol JG,ed.Proceedings of the Second International
Symposium on Menieres Disease.Amsterdam:Kuglerand Ghedini Publications, 170171.
Gibson WPR,Beagley MA.(1976) Transtympanic elec-trocochleography in the investigation of retrocochleardisorders.Rev Laryngol 97(Suppl.):507516.
Gibson WPR, Moffat DA,Ramsden RT. (1977) Clinicalelectrocochleography in the diagnosis and manage-ment of Menieres disorder.Audiology 16:389401.
Goin DW, Staller SJ, Asher DL, Mischke RE. (1982)Summating potential in Menieres disease.Laryngoscope 92:13811389.
Gulick WL, Gescheider GA, Frisina RD. (1989)Hearing: Physiological Acoustics, Neural Coding, andPsychoacoustics. New York: Oxford University Press.
Harris D, Dallos P. (1979) Forward masking of audi-tory nerve fiber responses. J Ne ur op hy si ol42:10831107.
Horner KC, Cazals Y-S. (1988) Independent fluctua-tions of the round-window summating potential andcom